Navegando por Data de Publicação, começando com "2025-10-24"
Agora exibindo 1 - 6 de 6
- Resultados por página
- Opções de Ordenação
listelement.badge.dso-typeItem, Segmentação de vasos sanguíneos com poucos dados via transferência de informação de forma(Universidade Federal de São Carlos, 2025-10-24) Galvão, Wesley Nogueira; Comin, Cesar Henrique; https://lattes.cnpq.br/9563440403120931; https://orcid.org/0000-0003-1207-4982; https://lattes.cnpq.br/3164765403744408; https://orcid.org/0009-0001-8545-3126; Traina, Agma Juci Machado; Almeida Junior, Jurandy Gomes de; https://lattes.cnpq.br/5136155977351408; https://lattes.cnpq.br/4495269939725770; https://orcid.org/0000-0003-4929-7258; https://orcid.org/0000-0002-4998-6996Blood vessel segmentation is an essential task in various areas of biomedicine, such as ophthalmology and neurology. However, this task poses a significant challenge due to the scarcity of large annotated datasets, which consequently makes the manual annotation of new samples for training efficient deep learning models a costly endeavor. Transfer learning techniques have shown promise in overcoming these limitations, allowing the reuse of representations learned in one domain to improve performance in another, even when few examples are available in the target domain. The central hypothesis indicates that leveraging a shape prior of vessel-like forms, such as their tubular and branching characteristics, can lead to more robust and data-efficient models. In this context, this dissertation investigates the transfer of shape representations from a synthetic domain to the segmentation of medical images in a few-shot regime. To this end, we introduce VessShape, a methodology for generating a large-scale synthetic dataset designed to instill a strong shape bias in segmentation models. VessShape images combine procedurally generated tubular geometries with a wide variety of textures, encouraging models to learn shape cues over appearance features. The models pre-trained with VessShape were then fine-tuned and evaluated on two real-world datasets from different domains. The results demonstrate that the approach achieves strong segmentation performance in few-shot scenarios, requiring only a small number of samples for fine-tuning. Additionally, the models demonstrate a significant zero-shot learning capability, proving able to segment vessels in unseen domains without any target-specific training. These results support that pre-training with a strong shape bias constitutes an effective strategy to overcome data scarcity and enhance generalization in blood vessel segmentation.listelement.badge.dso-typeItem, Estratégias quimiométricas utilizando espectroscopia vibracional e imagens digitais para análise de sal do Himalaia, açúcar mascavo e azeite de dendê(Universidade Federal de São Carlos, 2025-10-24) Melo, Welem de Oliveira Santiago de; Carneiro, Renato Lajarim; https://lattes.cnpq.br/8065852319463976; https://lattes.cnpq.br/4178212934295218; https://orcid.org/0000-0002-9271-8401; Pereira Filho, Edenir Rodrigues; Bogusz Junior, Stanislau; https://lattes.cnpq.br/3394181280355442; https://lattes.cnpq.br/9611719556320083; https://orcid.org/0000-0003-0608-0278Vibrational spectroscopy and digital imaging techniques, when combined with chemometric analysis, are widely applicable in food analysis. The modification of food matrices, whether through derivatization, partitioning, pre-concentration, digestion, drying, or other processes, can help reduce the presence of interfering substances in the samples, thereby improving the analytical performance of vibrational spectroscopy. In this work, 3 applications are proposed: 1) evaluation of commercial Himalayan salt sam-ples using digital imaging, Raman spectroscopy, and pre-extraction of poorly soluble compounds for LIBS analysis; 2) assessment of Raman spectroscopy for detecting adulteration in brown sugar, as well as the potential of FTIR spectroscopy for discriminating samples based on poorly soluble compounds extracted from them; 3) evaluation of commercial palm oil samples by Raman spectroscopy.listelement.badge.dso-typeItem, “O ontem - o hoje - o agora”: r(existencia) de mulheres negras através da amamentação(Universidade Federal de São Carlos, 2025-10-24) Stofel, Natália Sevilha; https://lattes.cnpq.br/2447361827796110; https://orcid.org/0000-0002-5928-3477listelement.badge.dso-typeItem, Microbioma ruminal ativo de Nelore: efeitos da dieta e interações com emissão de metano, eficiência alimentar(Universidade Federal de São Carlos, 2025-10-24) Silva, Juliana Virginio da; Andrade, Bruno Gabriel Nascimento; https://lattes.cnpq.br/0951322873579960; Regitano, Luciana Correia de Almeida; https://lattes.cnpq.br/9595338480545794; https://lattes.cnpq.br/9643767328433881; Regitano, Luciana Correia de Almeida; Nunes, Francis de Morais Franco; Pranchevicius, Maria Cristina da Silva; Cesar, Aline Silva Mello; Ibelli, Adriana Mércia Guaratini; https://lattes.cnpq.br/9595338480545794; https://lattes.cnpq.br/1974857490729279; https://lattes.cnpq.br/4772148921558977; https://lattes.cnpq.br/6423577693155656; https://lattes.cnpq.br/4729789236742545The growing global demand for beef, particularly from production systems in tropical regions, requires innovative strategies that enhance feed efficiency and reduce the environmental impacts of livestock. The ruminal microbiome plays a central role in the digestion of complex carbohydrates and the production of volatile fatty acids, which are the primary energy source for ruminants. In Brazil, the Nelore breed (Bos indicus) and its crossbreeds represent approximately 80% of the national herd. However, functional studies on the ruminal microbiome of this breed are still scarce. Advances in omics technologies now allow the investigation of how diet, the microbiome, and the host genome affect productive and environmental efficiency in beef cattle. Therefore, this study aimed to investigate the activity of the ruminal microbiome in a population of 52 Nelore cattle subjected to two distinct diets (conventional and byproducts) and to integrate, in an unprecedented manner, metagenomic, metatranscriptomic, and rumen wall microRNA data. To achieve this, we set the following objectives: Characterize the active functional profile of the ruminal microbiome in response to different diets using metatranscriptomic data; Assess the taxonomic and functional composition of the active ruminal microbiome by integrating metagenomic and metatranscriptomic data; Associate the active microbiome with methane emissions and feed efficiency; Investigate the interaction between the active microbiome and the host genome by integrating rumen wall microRNA expression with functional and taxonomic information of the microbiome. Metatranscriptomic analyses revealed key functional shifts in the rumen microbiome, despite relatively stable taxonomic composition across diets. Dietary components directly influenced microbial metabolism. In the conventional diet group, most expressed functions were associated with the metabolism of readily fermentable carbohydrates such as starch and mannose, present in corn and soybean meal. In contrast, animals receiving the by-product diet exhibited microbial activities linked to the degradation of complex carbohydrates, such as the cellulose found in citrus pulp. Additionally, the inclusion of by-products favored alternative microbial pathways that compete with methanogenesis, such as the reductive acetogenesis pathway (e.g., tetrahydrofolate synthase), which can redirect H₂ and CO₂ toward acetate production. Increased expression of folylpolyglutamate synthase/dihydropteroate synthase was also observed, potentially indicating methionine limitation and reducing the availability of essential substrates for methanogenesis, such as methanofuran and methanopterin. To further deepen our understanding of the rumen microbiome, we constructed a rumen prokaryotic gene catalog that revealed taxonomic and functional differences between the total metagenome (Nelore Ruminal Prokaryotic Microbiome Genes – NRPMG) and its transcriptionally active fraction (active-Nelore Ruminal Prokaryotic Microbiome Genes – aNRPMG). These findings indicate that fermentative processes in the rumen are shaped primarily by the functional activity of microbial communities, rather than their mere presence. We also identified enzymes and taxa associated with methane emissions and feed efficiency. Among them, Carbohydrate Esterase family 20 (CE20), Glycoside Hydrolase family 15 (GH15), and the genera Monoglobus and Halosimplex emerged as potential indicators or modulators of host performance traits. Finally, the analysis of host–microbiome interactions revealed co-expression patterns among rumen wall microRNAs, enzymes, and active microbial genera, suggesting meaningful host–microbiome interplay. These microRNAs may modulate the gene expression of enzymes or microorganisms involved in specific metabolic functions, including lipid metabolism and host health processes. Overall, the findings of this thesis reinforce that the rumen microbiome should be understood not only through its composition but, more importantly, through its functional activity. The integrative approach adopted here highlights how the interplay among microbiome, host, and diet shapes key metabolic processes and underscores microbial and molecular targets with potential to support strategies aimed at mitigating methane emissions and improving feed efficiency in cattle raised in tropical environments.listelement.badge.dso-typeItem, Apoio à família da pessoa com altas habilidades ou superdotação(Universidade Federal de São Carlos, 2025-10-24) Braz, Paula Paulino; Rangni, Rosemeire de Araújo; https://lattes.cnpq.br/6399149504309769; https://orcid.org/0000-0002-8752-9745; https://lattes.cnpq.br/7708298574440290; https://orcid.org/0000-0002-1571-5944; Sakaguti, Paula Mitsuyo Yamasaky; Pereira, Josilene Domingues Santos; Orlando, Rosimeire Maria; Pedro, Ketilin Mayra; https://lattes.cnpq.br/6605471166573621; https://lattes.cnpq.br/5147978033348152; https://lattes.cnpq.br/4254571422608459; https://lattes.cnpq.br/0991868156805558; https://orcid.org/0000-0002-8800-457X; https://orcid.org/0000-0001-5392-9006; https://orcid.org/0000-0002-0990-6146; https://orcid.org/0000-0002-1893-5002Family support for the development of the potential of gifted individuals is recommended; however, disruptions in family dynamics and lack of support can generate conflicts, insecurities, and feelings of disorientation on the part of parents when faced with the emotional, social, and academic demands of their children. The guiding question of this research was: how has support for the families of gifted individuals been provided in different regions of Brazil, according to parents or guardians? The main objective was to understand how support is provided to families for gifted individuals in different regions of Brazil. The specific objectives were to identify the main demands related to the organization of work focused on families, maping the existing support and assistance by Brazilian regions, and develop a guidance manual for supporting the families of gifted individuals. The study was descriptive with a multi-method approach, combining qualitative and quantitative procedures as complementary strategies for data analysis. Data collection was carried out using a questionnaire sent to parents or guardians of gifted individuals in all regions of Brazil via social media. Twenty-nine family members participated of the research. The results pointed to a) common challenges faced by families, such as difficulty accessing reliable information about giftedness, lack of institutional support networks, the coordination among educational, health, and social services; b) lack of systematic and intersectoral actions to welcome and guide families which reinforces the need for more effective support strategies; c) most participants had favorable socioeconomic conditions and a good level of education; d) there were discrepancies in the support provided to families in different regions of Brazil; and e) based on the results obtained, a guide was created for the families of gifted individuals which provides guidance for family members and guardians with the aim of contributing to the realization of the educational, social, and emotional rights for gifted individuals and their families. It is hoped that this research will guide actions that enable families to receive the attention they need to welcome their highly skilled or gifted children.listelement.badge.dso-typeItem, Efeitos dos exercícios físicos no manejo dos sintomas não motores de pessoas com a doença de Parkinson: revisões sistemáticas e ensaio clínico controlado(Universidade Federal de São Carlos, 2025-10-24) Costa, Valton da Silva; Gianlorenço, Anna Carolyna Lepesteur; https://lattes.cnpq.br/3861050459671690; https://orcid.org/0000-0002-7334-3835; Andrade, Larissa Pires de; https://lattes.cnpq.br/7334391500340646; https://orcid.org/0000-0002-8860-5818; https://lattes.cnpq.br/8403191957468940; https://orcid.org/0000-0002-2356-7523; Fiorin, Fernando da Silva; Russo, Thiago Luiz de; Navega, Flávia Roberta Faganello; Takahashi, Anielle Cristhine de Medeiros; https://lattes.cnpq.br/1783079128689681; https://lattes.cnpq.br/2755480908047791; https://lattes.cnpq.br/4080085921666959; https://lattes.cnpq.br/8336705133403404; https://orcid.org/0000-0002-6517-7160; https://orcid.org/0000-0002-5483-3097; https://orcid.org/0000-0002-5457-1154; https://orcid.org/0000-0003-1263-2833This thesis is the result of four years of research within the Graduate Program in Physical Therapy on the topic of the effects of physical exercise in the management of non-motor symptoms (NMS) of Parkinson’s disease (PD). The origin of these symptoms is multisystemic and involves aspects that affect mood, cognition, behavior, sleep, and autonomic and sensory functions. These symptoms, along with classic motor symptoms, negatively impact the quality of life of people with this condition. The objectives of the studies presented in this thesis were to investigate the clinical effects of physical exercise on these symptoms and identify the most effective exercise modalities. Furthermore, to investigate, for the first time, the effects of three types of upper limb (UL) physical exercises (resistance [RES], aerobic [AER], and task-specific training [TET]) and compare them to determine the most effective for NMS. The methodology of this thesis begins with systematic reviews and meta-analyses to achieve a comprehensive and up-to-date literature synthesis. In the experimental design, a randomized controlled clinical trial (RCT) was conducted with 37 individuals diagnosed with PD (Hoehn and Yahr stages 1-3). The aim was to compare the effects of different modalities of UL physical exercises, performed over 8 weeks, in relation to a control group receiving usual care without physical exercise. Outcomes were assessed using specific scales of NMS of PD, as well as standardized cognitive and neuropsychiatric assessment instruments. To analyze the RCT data, we adopted robust statistical models to compare the effects of the interventions (mixed-effects models, nonparametric models, and post hoc tests). Meta-analyses results revealed significant clinical effects of physical exercises on general NMS, general and executive cognitive function, depression, anxiety, fatigue, and sleep problems. Some modalities stood out, such as AER exercises for general NMS and depression, and mind-body exercises and RES for depression and anxiety. In the results of the RCT, betweengroup comparisons significantly demonstrated that UL physical exercises equally reduced overall NMS, cognitive dysfunction, and fatigue, though with different effect sizes. We observed that RES exercise was more effective in reducing symptoms of apathy and that TET was more effective in reducing symptoms of anxiety. Furthermore, within-group effects pointed to the clinical impact of RES exercise on general cognitive function, memory, visuospatial ability, depression, anxiety, sleep problems, and hallucinations; and of TET on executive function, depression, and hallucinations. Strengths and limitations of the methodology employed are discussed in this thesis. An important impact of this work is the need to raise awareness among healthcare professionals who care for people with PD about the existence and impact of NMS, as well as the possibility of using physical exercises to treat these symptoms. It is also expected that future clinical guidelines will begin to integrate this body of evidence on the use of physical exercises in the management of NMS, including UL exercises, into their recommendations. In this way, an important health need for people with PD will be met, contributing to improving functionality and health-related quality of life, and reducing the impacts on family members, caregivers, and the health system.